The assessment of insight in clinical psychiatry: a new scale I
Markova IS, Berrios GE. The assessment of insight in clinical psychiatry: a new scale. Acta Psychiatr Scand 1992: 86: 159-164. Difficulties surrounding the meaning of insight in psychiatry have led to neglect of its assessment in clinical practice. A study is described in which an insight scale was constructed and tested in a sample of 43 patients suffering from schizophrenia or depression. The results showed that the insight instrument was able to provide a qualitative and quantitative assessment of insight. Qualitative features brought out by the scale included patients’ perception of changes within themselves and within their environment, their recognition of being ill and their acknowledgment of needing help. Quantitative measurements showed that there was a significant improvement in the level of insight over the period of hospital admission. In addition, scores obtained on the insight scale were inversely correlated with the severity of the patients’ disorder, though these correlations occurred at different times in relation to the different disorders. We thus postulate that other factors (such as intellect, past experience, personality etc.) must be involved in modulating the expression of insight. Although the insight instrument needs refining, we suggest that it may be a useful tool in the further exploration of insight in clinical practice.
The assessment of insight in clinical psychiatry is part of the Mental State Examination; nevertheless, little is known on how it should be carried out or, indeed, on its value in respect to the patient and disorder. A reason for this seems to relate to the definition of insight (l),which remains neglected (2); and yet, the evaluation of insight is crucially dependent on a clear understanding of the term. In this article we ask three questions. Is it possible to evaluate insight clinically and identify its aspects and gradations? Do changes in insight occur over time? Are levels of insight related to the severity of the patient’s disorder? The answers to such questions lie in empirical work and a review of the existing literature has been carried out elsewhere (1). Studies concentrate more on the relationship between insight and other factors such as prognosis (3,4) and adherence to medication ( 5 , 6 ) than on its assessment (3,4). Insight tends to be considered as tantamount to the patient’s acceptance of illness and treatment (5); an exception is the work by McEvoy et al. (7, S), who devised an Insight and Treatment Attitude Questionnaire to compare the degrees and level of insight between voluntary and involuntary patients (7) and to relate levels of insight to severity of psychopathology (8). The questionnaire consists of 11 items concerning attitudes towards being mentally ill and requiring medication and hospital-
1
I. S. Markova’, G. E. Berries'
’
Department of Psychiatry, Fulbourn Hospital, Cambridge, * Department of Psychiatry, University of Cambridge, United Kingdom
Key words: insight: rating scale; schizophrenia; affective disorder
G.E. Berrios, Department of Psychiatry, University of Cambridge, Addenbrooke‘s Hospital (Level 4). Hills Road, Cambridge CB2 2QQ, United Kingdom
I
Accepted for publication March 7, 1992
ization and relating these to patients’ past and present circumstances and anticipation of future needs. Patients were invited to elaborate on their answers, which were then scored by a consensus of 3 judges; how such scores were decided, however, is not made clear. Using this scale, the authors found that voluntary patients with schizophrenia showed greater levels of insight than involuntary ones but that the degree of insight was not consistently related to severity. They thus postulated that levels of insight were not solely related to psychopathology. In contrast to previous work, the study attempted to measure insight not as a symptom that is present or absent but as a dynamic process that could be graded. Nevertheless, their method remains limited by their narrow definition of insight: the realization by the patient that he or she is mentally ill and requires treatment. David (2) has suggested another method of assessing insight in psychotic patients. His schedule appears to explore insight beyond the recognition of being ill and needing treatment, but also requires patients to identify the morbidity of their psychotic experience. Only a wider use of this scale will prove its value. Why has there been little progress in the clinical assessment of insight? Part of the reason might lie in the persistence of regarding insight as a symptom 159
Markova & Berrios
that is either present or absent, and defining it within narrow margins. For example, such major studies as the international pilot study of schizophrenia (9) refer to lack of insight as the most frequent symptom of acute schizophrenia. Although some psychiatric textbooks argue against this viewpoint (lo), they offer few practical alternatives. Yet it seems that it is the symptom model of insight that precludes more detailed assessment. We propose that the definition of insight be broadened. In a previous analysis (l), we suggested that insight be considered a subcategory of selfknowledge, which individuals hold not only about the disorder affecting them but also about how the disorder affects their interaction with the world. This is a complex notion and it would be unrealistic to expect that insight can be assessed fully, in its entirety, however, such a definition has the advantage of considering insight as a process, thereby allowing the longitudinal and dynamic aspects of insight to be examined. This article reports a study in which an insight scale was constructed and administered to 43 patients. The scale is still considered as a preliminary or exploratory tool to be modified by future research. Material and methods
Forty-three patients admitted consecutively to the acute general psychiatric wards in Fulbourn Hospital, Cambridge, were interviewed within 48 h of admission. At this stage we were concerned simply with testing the scale in patients in whom the diagnosis was straightforward, and in whom the severity of their disorder could be reliably ascertained. Because of difficulties in classifying and rating severity of personality disorders, these were excluded, as were disorders relating to substance abuse. Manic patients were also excluded because only 3 patients were seen during the study. The following protocol was completed: data sheet; the preliminary Insight Scale (Table 1); a semistructured interview to examine insight (available from authors); and scales to measure the severity of particular disorders, either - Hamilton Rating Scale for Depression (HRSD) (1 1) or - Brief Psychiatric Rating Scale (BPRS) (12). Of the 43 patients, 22 were reassessed using the same protocol within 48 h of discharge from hospital. Because of early or unplanned discharges, it was not possible to follow up every patient. The SPSSPCv.3.1 statistical package was used for data analysis. 160
Table 1. Insight Scale (preliminary) Instructions: please ring reply as required:
I have come into hospital for a rest. I have never been so ill in my life. Mental illness does not exist. I am here because I was asked to come. 5. My condition can be treated with medicines. 6. Why have you come to hospital? (you can ring more than one) a. My doctor asked me to. b. I am ill and need treatment. c. My wifdhusband might leave me if I didn't. d. I feel nervous but not ill. e. I was forced. 7. Should anyone be here instead of you? 8. To feel well I only need some advice and talking to. 9. I have been having some silly thoughts. 10. Nothing is the matter with me. 1 1, The mind cannot become ill, only the body. 12. I shall sue the hospital if I am not allowed out. 13. No one believes I am ill. 14. Something very strange is happening to me. 15. My neighbours are after me. 16. I feel my mind is going. 17. I know that my thoughts are silly but I cannot help it. 18. I cannot stop worrying about things. 19. The voices I hear are not in my mind. 20. Someone is controlling my mind. 21. All I need is to pull myself together. 22. I feel different from my normal self. 23. I am losing contact with my environment. 24. I am losing contact with myself. 25. I understand why I am in hospital. 26. I understand why other people think I should be in hospital. 27. I feel in control of my thoughts. 28. I feel in control of my feelings. 29. I could have prevented this situation, 30. I find it difficult to explain how I feel. 31. I want to know what is happening to me. 32. I want to know why I am feeling like this. 1. 2. 3, 4.
yes yes yes yes yes
no no no no no
don't know don't know don't know don't know don't know
yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes
no no no no no no no no no no no no no no no no no no no
don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know
yes yes yes yes yes yes yes
no no no no no no no
don't know don't know don't know don't know don't know don't know don't know
The Insight Scale
The Insight Scale (Table 1) has 32 items to be answered yes, no or don't know. The scale was administered by one observer (I.S.M.)but can also be self-rated. Hence, issues of interrater reliability do not arise. Items were chosen on a face validity method, i.e. by breaking up the concept of insight into components thought to represent aspects of self-knowledge relevant to the patient's illness. Questions pertaining to the following areas were included: hospitalization (1, 4, 6a, 7, 21, 25, 26); mental illness in general (3, 11); perception of being ill (2, 5, 6b, 10, 13); changes in the self (6d, 8,9,14, 16, 17, 18,21,22, 23, 24, 30); control over the situation (20, 27, 28, 29); perception of the environment (6c, 15,20,23,26); and wanting to understand one's situation (31, 32).
Assessment of insight in clinical psychiatry Scoring
To facilitate analysis, a score of 2 was given for a positive response and 1 for a negative response, and items were subdivided into those that a) if answered positively would indicate greater (positive) insight ( 5 , 6b, 8, 9, 13, 14, 16, 17, 18, 22, 23, 24, 26, 32)(group A), and b) if answered positively would indicate less (negative) insight (1, 3, 7, 10, 11, 15, 20) (group B). T o determine the direction of insight in respect to individual items, the following assumptions were made: Items relating to general attitudes towards hospitalization, mental illness and medication (categories 1 and 2) mainly reflect the individual’s cultural background. Positive insight was scored when the patient acknowledged that mental illness existed and could be treated with medication. Positive insight was also scored when individuals recognized they were ill (and more specifically mentally ill) and in need of treatment (category 3). It was assumed that when individuals become mentally disordered a number of cognitive and experiential changes occur, affecting perception of self, of environment and of the interaction between the two. The items in categories 4 , 5 , 6 and 7 relate to the awareness of such changes. Thus, positive insight was scored when the patient acknowledged a difference within him- or herself and with his or her interaction with the outside world. ltems found to be ambiguous in respect to the direction of insight were discarded (2, 4, 6a, 6c, 6d, 6e, 12, 19,21,27,28, 29) before analysis took place. With regard to “don’t know” responses, these were discarded when no firm conclusions could be drawn concerning the direction of insight. Semistructured interview
This instrument was used in conjunction with the scale to allow further exploration of insight and to bring out any other aspects of insight which might be useful in the further development of the scale. Information from this interview was used to check on the veracity of answers obtained to the insight scale, but analysis of the qualitative data collected by this instrument is not included in this article. Results
The demographic data and diagnoses for the sample are shown in Table 2. The results are summarized in 2 parts. The first deals with the analysis of the preliminary Insight
Table 2. Demographic data and diagnoses Interview 1
Interview 2**
Numbers of patients
43
22
Sex
men women
19 24
11 11
Age (years)
mean range
43 22-68
46 22-68
19 (BPRSa range 33-60)
12 (BPRSa range 33-60, BPRSb range 17-33]
Diagnosis schizophonia (DSM-Ill-R) Major depression
13 (HAMa range 21-44)
5 (HAMa range 33-42, HAMb range 2-1 1)
Dysthymia
7 (HAMa range 24-33]
4 (HAMa range 24-33, HAMb range 3-1 1I
Bipolar depression
4 (HAMa range 23-30)
1 (HAMa 27, HAMb 12)
a=score at interview 1, b=score at interview 2. ** It was not possible to follow up 2 1 patients: 1 1 were discharged before we could be notified: 6 took their own discharge; 2 refused to cooperate with the 2nd interview; 1 was transferred to another hospital; 1 died. The attrition rate, however, should not affect the results concerning the structure of the scale, as the reason for the 2nd interview was to look at the sensitivity to change of the Insight Scale.
Scale based on 43 patients; the second with the analysis of the results based on the 22 patients interviewed twice. Analysis of the preliminary Insight Scale
Questions belonging to groups A (positive insight) and B (negative insight) were analysed separately by exploratory principal component analysis (factor analysis). Group A yielded 5 factors, which accounted for 63.5% of the variance (lowest eigenvalue 1.21): Factor 1 (23% of variance). This factor contains items concerning both perception of being ill and awareness of changes happening within the self. There is also the acknowledgment that other people recognize a change in the patient. Factor 2 (12.4 % of variance). These items are specifically concerned with the patients’ recognition of thought processes being affected by the illness. Factor 3 (10.1 % of variance). This factor is concerned with patients’ awareness of changes within themselves and the reflection of such changes in their interaction with their environment. There also appears to be a need for self-understanding. Factor 4 (9.3% of variance). These items reflect general awareness of changes within the self and the resulting difficulties in functioning. Factor 5 (8.7% of variance). This factor is concerned with the recognition of self-change and the 161
Markova & Berrios need for help. Items also indicate an expectation that the situation is treatable. The common theme underlying all these factors is the patient’s recognition of a change in the self. Group B yielded 3 factors and accounted for 63.6% of the variance (lowest eigenvalue 1.13): Factor 1 (28.3% of variance). Items in this factor relate to perception of being ill and also to attitudes towards mental illness in general. Factor 2 (19.1 % of variance). This factor is concerned with perception of changes in the environment and may indicate delusional thinking. Factor 3 (16.2 % of variance). This factor is concerned with attitudes towards hospitalization. Reliability
Table 3 Comparison of global scores obtained for group A and group B items between 1st and 2nd interviews Group A (insight-positive)
Interview 1 Interview 2
Group B* (insight-negative)
Mean
SD
Mean
SD
18.2 20.3
3.5 2.3
11.5 12.6
1.8 1.2
* For facility, the scoring of insight-negative or group B items has been reversed: the greater the score, the greater the level of insight.
a greater degree of severity of depression around the time of discharge had lower levels of insight at that time.
Items in groups A and B were analysed for internal consistency using Cronbach’s alpha coefficient. This is an estimate of test reliability based on item intercorrelations (13). Group A yielded a standardized alpha=0.71 and group B an alpha=0.55 at this stage of the study, these alphas should be considered as adequate. No item proved to be redundant (i.e. its rejection increasing significantly the alpha value) in either group.
At first interview, patients with schizophrenia (19) had higher positive insight scores than patients with depression (24) (Mann-Whitney U-test, P c 0.02). No significant differences were found between these two diagnostic groups in regard to negative insight scores.
Changes in insight
Discussion
For the 22 patients seen twice, overall scores for items in groups A and B were calculated separately for interview 1 and interview 2. Then a comparison was made of scores at first and at second interview, using a two-tailed t-test for related samples (Table 3). The data in Table 3 show a significant change in both positive ( t = 3.54, P
Markova IS, Berrios GE. The assessment of insight in clinical psychiatry: a new scale. Acta Psychiatr Scand 1992: 86: 159-164. Difficulties surrounding the meaning of insight in psychiatry have led to neglect of its assessment in clinical practice. A study is described in which an insight scale was constructed and tested in a sample of 43 patients suffering from schizophrenia or depression. The results showed that the insight instrument was able to provide a qualitative and quantitative assessment of insight. Qualitative features brought out by the scale included patients’ perception of changes within themselves and within their environment, their recognition of being ill and their acknowledgment of needing help. Quantitative measurements showed that there was a significant improvement in the level of insight over the period of hospital admission. In addition, scores obtained on the insight scale were inversely correlated with the severity of the patients’ disorder, though these correlations occurred at different times in relation to the different disorders. We thus postulate that other factors (such as intellect, past experience, personality etc.) must be involved in modulating the expression of insight. Although the insight instrument needs refining, we suggest that it may be a useful tool in the further exploration of insight in clinical practice.
The assessment of insight in clinical psychiatry is part of the Mental State Examination; nevertheless, little is known on how it should be carried out or, indeed, on its value in respect to the patient and disorder. A reason for this seems to relate to the definition of insight (l),which remains neglected (2); and yet, the evaluation of insight is crucially dependent on a clear understanding of the term. In this article we ask three questions. Is it possible to evaluate insight clinically and identify its aspects and gradations? Do changes in insight occur over time? Are levels of insight related to the severity of the patient’s disorder? The answers to such questions lie in empirical work and a review of the existing literature has been carried out elsewhere (1). Studies concentrate more on the relationship between insight and other factors such as prognosis (3,4) and adherence to medication ( 5 , 6 ) than on its assessment (3,4). Insight tends to be considered as tantamount to the patient’s acceptance of illness and treatment (5); an exception is the work by McEvoy et al. (7, S), who devised an Insight and Treatment Attitude Questionnaire to compare the degrees and level of insight between voluntary and involuntary patients (7) and to relate levels of insight to severity of psychopathology (8). The questionnaire consists of 11 items concerning attitudes towards being mentally ill and requiring medication and hospital-
1
I. S. Markova’, G. E. Berries'
’
Department of Psychiatry, Fulbourn Hospital, Cambridge, * Department of Psychiatry, University of Cambridge, United Kingdom
Key words: insight: rating scale; schizophrenia; affective disorder
G.E. Berrios, Department of Psychiatry, University of Cambridge, Addenbrooke‘s Hospital (Level 4). Hills Road, Cambridge CB2 2QQ, United Kingdom
I
Accepted for publication March 7, 1992
ization and relating these to patients’ past and present circumstances and anticipation of future needs. Patients were invited to elaborate on their answers, which were then scored by a consensus of 3 judges; how such scores were decided, however, is not made clear. Using this scale, the authors found that voluntary patients with schizophrenia showed greater levels of insight than involuntary ones but that the degree of insight was not consistently related to severity. They thus postulated that levels of insight were not solely related to psychopathology. In contrast to previous work, the study attempted to measure insight not as a symptom that is present or absent but as a dynamic process that could be graded. Nevertheless, their method remains limited by their narrow definition of insight: the realization by the patient that he or she is mentally ill and requires treatment. David (2) has suggested another method of assessing insight in psychotic patients. His schedule appears to explore insight beyond the recognition of being ill and needing treatment, but also requires patients to identify the morbidity of their psychotic experience. Only a wider use of this scale will prove its value. Why has there been little progress in the clinical assessment of insight? Part of the reason might lie in the persistence of regarding insight as a symptom 159
Markova & Berrios
that is either present or absent, and defining it within narrow margins. For example, such major studies as the international pilot study of schizophrenia (9) refer to lack of insight as the most frequent symptom of acute schizophrenia. Although some psychiatric textbooks argue against this viewpoint (lo), they offer few practical alternatives. Yet it seems that it is the symptom model of insight that precludes more detailed assessment. We propose that the definition of insight be broadened. In a previous analysis (l), we suggested that insight be considered a subcategory of selfknowledge, which individuals hold not only about the disorder affecting them but also about how the disorder affects their interaction with the world. This is a complex notion and it would be unrealistic to expect that insight can be assessed fully, in its entirety, however, such a definition has the advantage of considering insight as a process, thereby allowing the longitudinal and dynamic aspects of insight to be examined. This article reports a study in which an insight scale was constructed and administered to 43 patients. The scale is still considered as a preliminary or exploratory tool to be modified by future research. Material and methods
Forty-three patients admitted consecutively to the acute general psychiatric wards in Fulbourn Hospital, Cambridge, were interviewed within 48 h of admission. At this stage we were concerned simply with testing the scale in patients in whom the diagnosis was straightforward, and in whom the severity of their disorder could be reliably ascertained. Because of difficulties in classifying and rating severity of personality disorders, these were excluded, as were disorders relating to substance abuse. Manic patients were also excluded because only 3 patients were seen during the study. The following protocol was completed: data sheet; the preliminary Insight Scale (Table 1); a semistructured interview to examine insight (available from authors); and scales to measure the severity of particular disorders, either - Hamilton Rating Scale for Depression (HRSD) (1 1) or - Brief Psychiatric Rating Scale (BPRS) (12). Of the 43 patients, 22 were reassessed using the same protocol within 48 h of discharge from hospital. Because of early or unplanned discharges, it was not possible to follow up every patient. The SPSSPCv.3.1 statistical package was used for data analysis. 160
Table 1. Insight Scale (preliminary) Instructions: please ring reply as required:
I have come into hospital for a rest. I have never been so ill in my life. Mental illness does not exist. I am here because I was asked to come. 5. My condition can be treated with medicines. 6. Why have you come to hospital? (you can ring more than one) a. My doctor asked me to. b. I am ill and need treatment. c. My wifdhusband might leave me if I didn't. d. I feel nervous but not ill. e. I was forced. 7. Should anyone be here instead of you? 8. To feel well I only need some advice and talking to. 9. I have been having some silly thoughts. 10. Nothing is the matter with me. 1 1, The mind cannot become ill, only the body. 12. I shall sue the hospital if I am not allowed out. 13. No one believes I am ill. 14. Something very strange is happening to me. 15. My neighbours are after me. 16. I feel my mind is going. 17. I know that my thoughts are silly but I cannot help it. 18. I cannot stop worrying about things. 19. The voices I hear are not in my mind. 20. Someone is controlling my mind. 21. All I need is to pull myself together. 22. I feel different from my normal self. 23. I am losing contact with my environment. 24. I am losing contact with myself. 25. I understand why I am in hospital. 26. I understand why other people think I should be in hospital. 27. I feel in control of my thoughts. 28. I feel in control of my feelings. 29. I could have prevented this situation, 30. I find it difficult to explain how I feel. 31. I want to know what is happening to me. 32. I want to know why I am feeling like this. 1. 2. 3, 4.
yes yes yes yes yes
no no no no no
don't know don't know don't know don't know don't know
yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes
no no no no no no no no no no no no no no no no no no no
don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know don't know
yes yes yes yes yes yes yes
no no no no no no no
don't know don't know don't know don't know don't know don't know don't know
The Insight Scale
The Insight Scale (Table 1) has 32 items to be answered yes, no or don't know. The scale was administered by one observer (I.S.M.)but can also be self-rated. Hence, issues of interrater reliability do not arise. Items were chosen on a face validity method, i.e. by breaking up the concept of insight into components thought to represent aspects of self-knowledge relevant to the patient's illness. Questions pertaining to the following areas were included: hospitalization (1, 4, 6a, 7, 21, 25, 26); mental illness in general (3, 11); perception of being ill (2, 5, 6b, 10, 13); changes in the self (6d, 8,9,14, 16, 17, 18,21,22, 23, 24, 30); control over the situation (20, 27, 28, 29); perception of the environment (6c, 15,20,23,26); and wanting to understand one's situation (31, 32).
Assessment of insight in clinical psychiatry Scoring
To facilitate analysis, a score of 2 was given for a positive response and 1 for a negative response, and items were subdivided into those that a) if answered positively would indicate greater (positive) insight ( 5 , 6b, 8, 9, 13, 14, 16, 17, 18, 22, 23, 24, 26, 32)(group A), and b) if answered positively would indicate less (negative) insight (1, 3, 7, 10, 11, 15, 20) (group B). T o determine the direction of insight in respect to individual items, the following assumptions were made: Items relating to general attitudes towards hospitalization, mental illness and medication (categories 1 and 2) mainly reflect the individual’s cultural background. Positive insight was scored when the patient acknowledged that mental illness existed and could be treated with medication. Positive insight was also scored when individuals recognized they were ill (and more specifically mentally ill) and in need of treatment (category 3). It was assumed that when individuals become mentally disordered a number of cognitive and experiential changes occur, affecting perception of self, of environment and of the interaction between the two. The items in categories 4 , 5 , 6 and 7 relate to the awareness of such changes. Thus, positive insight was scored when the patient acknowledged a difference within him- or herself and with his or her interaction with the outside world. ltems found to be ambiguous in respect to the direction of insight were discarded (2, 4, 6a, 6c, 6d, 6e, 12, 19,21,27,28, 29) before analysis took place. With regard to “don’t know” responses, these were discarded when no firm conclusions could be drawn concerning the direction of insight. Semistructured interview
This instrument was used in conjunction with the scale to allow further exploration of insight and to bring out any other aspects of insight which might be useful in the further development of the scale. Information from this interview was used to check on the veracity of answers obtained to the insight scale, but analysis of the qualitative data collected by this instrument is not included in this article. Results
The demographic data and diagnoses for the sample are shown in Table 2. The results are summarized in 2 parts. The first deals with the analysis of the preliminary Insight
Table 2. Demographic data and diagnoses Interview 1
Interview 2**
Numbers of patients
43
22
Sex
men women
19 24
11 11
Age (years)
mean range
43 22-68
46 22-68
19 (BPRSa range 33-60)
12 (BPRSa range 33-60, BPRSb range 17-33]
Diagnosis schizophonia (DSM-Ill-R) Major depression
13 (HAMa range 21-44)
5 (HAMa range 33-42, HAMb range 2-1 1)
Dysthymia
7 (HAMa range 24-33]
4 (HAMa range 24-33, HAMb range 3-1 1I
Bipolar depression
4 (HAMa range 23-30)
1 (HAMa 27, HAMb 12)
a=score at interview 1, b=score at interview 2. ** It was not possible to follow up 2 1 patients: 1 1 were discharged before we could be notified: 6 took their own discharge; 2 refused to cooperate with the 2nd interview; 1 was transferred to another hospital; 1 died. The attrition rate, however, should not affect the results concerning the structure of the scale, as the reason for the 2nd interview was to look at the sensitivity to change of the Insight Scale.
Scale based on 43 patients; the second with the analysis of the results based on the 22 patients interviewed twice. Analysis of the preliminary Insight Scale
Questions belonging to groups A (positive insight) and B (negative insight) were analysed separately by exploratory principal component analysis (factor analysis). Group A yielded 5 factors, which accounted for 63.5% of the variance (lowest eigenvalue 1.21): Factor 1 (23% of variance). This factor contains items concerning both perception of being ill and awareness of changes happening within the self. There is also the acknowledgment that other people recognize a change in the patient. Factor 2 (12.4 % of variance). These items are specifically concerned with the patients’ recognition of thought processes being affected by the illness. Factor 3 (10.1 % of variance). This factor is concerned with patients’ awareness of changes within themselves and the reflection of such changes in their interaction with their environment. There also appears to be a need for self-understanding. Factor 4 (9.3% of variance). These items reflect general awareness of changes within the self and the resulting difficulties in functioning. Factor 5 (8.7% of variance). This factor is concerned with the recognition of self-change and the 161
Markova & Berrios need for help. Items also indicate an expectation that the situation is treatable. The common theme underlying all these factors is the patient’s recognition of a change in the self. Group B yielded 3 factors and accounted for 63.6% of the variance (lowest eigenvalue 1.13): Factor 1 (28.3% of variance). Items in this factor relate to perception of being ill and also to attitudes towards mental illness in general. Factor 2 (19.1 % of variance). This factor is concerned with perception of changes in the environment and may indicate delusional thinking. Factor 3 (16.2 % of variance). This factor is concerned with attitudes towards hospitalization. Reliability
Table 3 Comparison of global scores obtained for group A and group B items between 1st and 2nd interviews Group A (insight-positive)
Interview 1 Interview 2
Group B* (insight-negative)
Mean
SD
Mean
SD
18.2 20.3
3.5 2.3
11.5 12.6
1.8 1.2
* For facility, the scoring of insight-negative or group B items has been reversed: the greater the score, the greater the level of insight.
a greater degree of severity of depression around the time of discharge had lower levels of insight at that time.
Items in groups A and B were analysed for internal consistency using Cronbach’s alpha coefficient. This is an estimate of test reliability based on item intercorrelations (13). Group A yielded a standardized alpha=0.71 and group B an alpha=0.55 at this stage of the study, these alphas should be considered as adequate. No item proved to be redundant (i.e. its rejection increasing significantly the alpha value) in either group.
At first interview, patients with schizophrenia (19) had higher positive insight scores than patients with depression (24) (Mann-Whitney U-test, P c 0.02). No significant differences were found between these two diagnostic groups in regard to negative insight scores.
Changes in insight
Discussion
For the 22 patients seen twice, overall scores for items in groups A and B were calculated separately for interview 1 and interview 2. Then a comparison was made of scores at first and at second interview, using a two-tailed t-test for related samples (Table 3). The data in Table 3 show a significant change in both positive ( t = 3.54, P
